ArticlePDF Available

INVESTIGATING ECOLOGICAL AND PHYLOGENETIC CONSTRAINTS IN HIPPOPOTAMIDAE SKULL SHAPE

Authors:

Abstract and Figures

Hippopotamidae are a group of large-sized mammals of interest for testing evolutionary traits in time and space. Variation in skull shape within Hippopotamidae is here investigated by means of shape analysis (Ge-ometric Morphometrics) and modern statistical approaches. Two-dimensional shape analysis is applied to dorsal and lateral views of extant and extinct Hippopotamidae species sufficiently preserved to allow their morphology to be captured by landmark and semi-landmark digitization. The results show that Hippopotamus gorgops and H. antiquus display similar shapes, while Hexaprotodon palaeindicus falls within the morphospace occupied by H. amphibius, suggesting similar morphology. The cranial shape of the Sicilian hippopotamus (H. pentlandi) still resembles that of H. amphibius in lateral view, suggesting that adaptation to the insular domain was yet not fully attained. Madagascan hippopotamu-ses (H. madagascariensis and H. lemerlei) are close to the pygmy hippo, Choeropsis liberiensis, in PC1 values; nevertheless, the cranial shape of the Madagascan hippos seems not to be closely related to the cranial shape of C. liberiensis. Despite the morphological convergences within the group, while cranial shape in Hippopotamidae is phylogenetically structured, this does not hold for size. Although further investigations are needed to test the influence of ecological and palaeo-ecological parameters on the general shape to provide additional information for understanding Hippopotamidae evolution and adaptation, the present study provides an insight into the evolutionary framework of Hippopotamidae.
Landmark (red points) configurations of Hippopotamidae skulls in lateral (A) and dorsal (B) views. A: 1, posterior tip of the nuchal crest; 9, contact beween the orbit and the skull roof; 10, tip of the dorsal border orbit; 14, anterior border of the orbit; 18, tip of the ventral border of the orbit; 19, contact between the dorsal borde of the zygomatic arch and the posterior border of the orbit; 23, lowermost tip of the dorsal border of the zygomatic arch; 28, dorsal-posterior tip of the zygomatic arch; 29, lowermost tip of the ventral border of the zygomatic arch; 38, contact between the zygomatic arch and the maxilla; 39, posterior tip of the dental row; 40, orthogonal projection, relative to the horizontal axis identified by landmarks 39-40, of point 41 at the base of the dental row; 41, posterior border of the infraorbital foramen; 42, orthogonal projection, relative to the horizontal axis identified by landmarks 39-40, of point 41 at the cranial roof; 47, contact between the cranial roof and the anterior border of the orbit. Semilandmarks (white points): 2-8, 11-13, 15-17, 20-22, 24-27, 30-37, 43-46. B: 1, middle point of the posterior border of the nuchal crest; 2, anterior tip of the frontal-parietal crest; 3, orthogonal projection, relative to the horizontal axis identified by landmarks 1-2, of the point 4 in the middle of the cranial roof; 4, inflexion point on the dorsal view of right maxilla; 16, posterior-lateral tip of the nuchal crest; 17, posterior tip of the zygomatic cavity; 21, posterior tip of the dorsal border of the orbit; 22, anterior tip of the dorsal border of the orbit. Semilandmarks (white points): 5-15, 18-20.
… 
Content may be subject to copyright.
INVESTIGATING ECOLOGICAL AND PHYLOGENETIC CONSTRAINTS IN
HIPPOPOTAMIDAE SKULL SHAPE
LUCA PANDOLFI1*, ROBERTA MARTINO2, LORENZO ROOK1 & PAOLO PIRAS3
*Corresponding author. 1Dipartimento di Scienze della Terra, Paleo[Fab]Lab, Università degli Studi di Firenze, Via G. La Pira 4, Firenze, 50121,
Italy. E-mail: luca.pandol@uni.it; lorenzo.rook@uni.it
2Dipartimento di Scienze della Terra, Sapienza Università di Roma, Laurea Magistrale in Scienze Geologiche, Piazzale Aldo Moro 5, Rome,
I-00185, Italy. E-mail: roberta.aska@gmail.com
3Dipartimento di Scienze Cardiovascolari, Respiratorie, Nefrologiche, Anestesiologiche e Geriatriche, Sapienza Università di Roma, Viale del
Policlinico 155, Rome, 00161, Italy. E-mail: ppiras@uniroma3.it
To cite this article: Pandol L., Martino R., Rook L. & Piras P. (2020) - Investigating ecological and phylogenetic constraints in Hippopotamidae
skull shape. Riv. It. Paleontol. Strat., 126(1): 37-49.
Rivista Italiana di Paleontologia e Stratigraa
(Research in Paleontology and Stratigraphy)
vol. 126(1): 37-49. March 2020
Abstract. Hippopotamidae are a group of large-sized mammals of interest for testing evolutionary traits in
time and space. Variation in skull shape within Hippopotamidae is here investigated by means of shape analysis (Ge-
ometric Morphometrics) and modern statistical approaches. Two-dimensional shape analysis is applied to dorsal and
lateral views of extant and extinct Hippopotamidae species sufciently preserved to allow their morphology to be
captured by landmark and semi-landmark digitization. The results show that Hippopotamus gorgops and H. antiquus di-
splay similar shapes, while Hexaprotodon palaeindicus falls within the morphospace occupied by H. amphibius, suggesting
similar morphology. The cranial shape of the Sicilian hippopotamus (H. pentlandi) still resembles that of H. amphibius
in lateral view, suggesting that adaptation to the insular domain was yet not fully attained. Madagascan hippopotamu-
ses (H. madagascariensis and H. lemerlei) are close to the pygmy hippo, Choeropsis liberiensis, in PC1 values; nevertheless, the
cranial shape of the Madagascan hippos seems not to be closely related to the cranial shape of C. liberiensis. Despite the
morphological convergences within the group, while cranial shape in Hippopotamidae is phylogenetically structured,
this does not hold for size. Although further investigations are needed to test the inuence of ecological and palaeo-
ecological parameters on the general shape to provide additional information for understanding Hippopotamidae
evolution and adaptation, the present study provides an insight into the evolutionary framework of Hippopotamidae.
Received: March 26, 2019; accepted: July 31, 2019
Keywords: Geometric Morphometrics; cranial shape; phylogenetic signal; fossil and extant Hippopotamidae.
IntroductIon
Hippopotamidae are a group of large-sized
mammals that allow observation of variation
in morphological traits in response to different
environmental conditions (aquatic to terrestrial).
Specically, the development of the orbits extending
above the cranial roof has been interpreted as an
important semiaquatic adaptation within the group,
especially in the Hippopotamus and Hexaprotodon
lineages (Boisserie 2005, 2007; Boisserie et al. 2011).
The cranial morphology of the extant pygmy-hippo,
Choeropsis liberiensis, with maintenance of terrestrial
characters such as orbits below the cranial roof,
has been considered as a model for several extinct
species. The convergent appearance of terrestrial
characters, related to a dietary shift and dwarsm,
occurred within Hippopotamus-lineages during the
Pleistocene, in particular after colonization of
islands (e.g., Crete, Malta, Sicily and Madagascar)
(Sondaar 1977; Houtemaker & Sondaar 1979;
Stuenes 1989; Caloi & Palombo 1994; Spaan 1996).
Pandol L., Martino R., Rook L. & Piras P.
38
A selection of some insular hippo’s spe-
cies (H. lemerlei, H. madagascariensis and Phanourios
minor) from Pleistocene and Holocene of Cyprus
and Madagascar was recently the subject of a study
aimed at comparing their shape change due to insu-
larity with those occurring in Elephantidae (van der
Geer et al. 2018). Here we focus on a broader taxo-
nomic and temporal sampling of Hippopotamidae
solely, from Miocene to Recent, by including both
insular and non-insular species.
Despite several studies focused on the cra-
nial morphology of Hippopotamidae, the variation
in skull shape within this group has seldom been
investigated by means of shape analysis (Geo-
metric Morphometrics) and modern statistical ap-
proaches in order to quantify differences and de-
gree of homoplasy among the species, and in order
to test the inuence of phylogenetic and ecological
signals on the variation in skull shape. This paper
aims to investigate these aspects by means of two-
dimensional shape analysis applied to dorsal and
lateral views of extant and extinct Hippopotamidae
species sufciently well preserved in order to al-
low their morphology to be captured by landmark
and semi-landmark digitization in lateral and dorsal
views.
MaterIal and Methods
Material
We collected images of 49 skulls of hippo-
potamids from both original photos and published
pictures in lateral and dorsal views (Supplementary
Material). Eleven hippopotamid species are repre-
sented in the sample (two extant and nine extinct
species). In order to eliminate inter-observer error,
the same operator (L.P.) digitized the entire land-
mark dataset. Hippopotamid species represented
by at least one complete skull are recorded in the
sample. Fossil material is poorly represented by
complete skulls due to the difculties of preserva-
tion. As sex is directly observable in extant species
only, and as we are dealing with taxonomic diversity
spanning from the Miocene to Recent and among
different genera, we did not include sex as a vari-
able in our analyses. In addition, we excluded mor-
phological features strongly inuenced by sexual
dimorphism (tusks size and jowl area; Stevenson-
Hamilton 1912; Coughlin & Fish 2009) from the
shape conguration. The specimen list and the
number of individuals for each species, as well as
the list of institutions and references from which
the images used for shape acquisition were collect-
ed are reported in Supplementary Material.
Geometric Morphometrics
We adopted Geometric Morphometrics as
our shape analysis tool in order to analyze mor-
phological variation. GM is demonstrated to be an
effective method for analysis of anatomical varia-
tion and disparity in both extinct and extant taxa
(Piras et al. 2009, 2010, 2014; Maiorino et al. 2013;
Pandol & Maiorino 2016; Pandol et al. 2019).
We digitized 15 landmarks and 32 semi-land-
marks from photographs in lateral view (Fig. 1A)
and 9 landmarks and 13 semi-landmarks in dorsal
view (Fig. 1B), on each specimen using the tpsDig2
v2.17 software (Rohlf 2013). Given that the ros-
tral region is often damaged, or enterely lacking,
in fossil specimens we chose a shape congura-
tion allowing inclusion of the maximum number
of extinct species, without losing relevant mor-
phological signal but excluding incomplete or se-
verely damaged skulls. Semi-landmarks were used
to capture the morphology of complex outlines
where anatomical homology is difcult to recog-
nize. Semi-landmarks assume that curves or con-
tours are homologous among specimens (Adams et
al. 2004; Perez et al. 2006). Generalized Procrustes
Analysis (GPA) (Bookstein 1986; Goodall 1991),
implemented in the procSym( ) function from the
R-package “Morpho” (Schlager 2014), was used
to rotate, translate and scale landmark congura-
tions to unit centroid size (CS = the square root
of the sum of squared distances of the landmarks
from their centroid) (Bookstein 1991). We used the
minimization of bending energy during the slid-
ing of semi-landmarks (cfr. Gunz & Mitteroecker
2013). After GPA, a Principal Components Analy-
sis (PCA) was performed in order to visualize or-
thogonal axes of morphological variation. The Un-
weighted Pair Group Method with Arithmetic mean
(UPGMA) algorithm was used on the per-species
averaged Procrustes distances to assess similarities
among taxa. The results are dendrograms of mor-
phological similarities (skulls in lateral and dorsal
views) among species included in the sample. In
order to visualize shape changes in ordination plots
we choose to use the method described in Márquez
Hippopotamidae skull shape 39
et al. (2012). There it was suggested that a useful
way to visualize local, innitesimal variation within
a deformation grid is to use the Jacobian (J) of the
Thin Plate Spline interpolation function. This mea-
sures the rate of shape deformation at any point
along all directions simultaneously. As J contains
the rst partial derivatives of the TPS, the afne
component, which is a rst-order polynomial, be-
comes a constant and for this reason J captures in-
formation as localized variation in the non-afne
component of the deformation. In 2D J is a 2x2
matrix that can be evaluated at any point within a
body. The logarithm of its determinant represents
change in the area in the region about the interpo-
lation point. Values < 0 indicate that, with respect
to the source (here the sample’ consensus), the tar-
get (here the PC’s extremes) experiences a reduc-
tion in the local area, while values > 0 indicate an
enlargement.
We tested for signicant phylogenetic signal
in size (CS) using the function phylosig( ) in the
‘phytools’ R package (Revell 2012). The degree of
phylogenetic signal in shape data for a given phy-
logenetic tree was quantied using the geomorph
Fig. 1 - Landmark (red points) congurations of Hippopotamidae skulls in lateral (A) and dorsal (B) views. A: 1, posterior tip of the nuchal
crest; 9, contact beween the orbit and the skull roof; 10, tip of the dorsal border orbit; 14, anterior border of the orbit; 18, tip of the
ventral border of the orbit; 19, contact between the dorsal borde of the zygomatic arch and the posterior border of the orbit; 23,
lowermost tip of the dorsal border of the zygomatic arch; 28, dorsal-posterior tip of the zygomatic arch; 29, lowermost tip of the
ventral border of the zygomatic arch; 38, contact between the zygomatic arch and the maxilla; 39, posterior tip of the dental row;
40, orthogonal projection, relative to the horizontal axis identied by landmarks 39-40, of point 41 at the base of the dental row; 41,
posterior border of the infraorbital foramen; 42, orthogonal projection, relative to the horizontal axis identied by landmarks 39-
40, of point 41 at the cranial roof; 47, contact between the cranial roof and the anterior border of the orbit. Semilandmarks (white
points): 2-8, 11-13, 15-17, 20-22, 24-27, 30-37, 43-46. B: 1, middle point of the posterior border of the nuchal crest; 2, anterior tip
of the frontal-parietal crest; 3, orthogonal projection, relative to the horizontal axis identied by landmarks 1-2, of the point 4 in the
middle of the cranial roof; 4, inexion point on the dorsal view of right maxilla; 16, posterior-lateral tip of the nuchal crest; 17, pos-
terior tip of the zygomatic cavity; 21, posterior tip of the dorsal border of the orbit; 22, anterior tip of the dorsal border of the orbit.
Semilandmarks (white points): 5-15, 18-20.
Pandol L., Martino R., Rook L. & Piras P.
40
function ‘physignal’ (Adams et al. 2019). The func-
tion ‘phylo.heatmap’ has been used to create a
multivariate phylogenetic heatmap (Revell 2012).
We used phylogenetic generalized least squares
(PGLS) regressions from package ‘geomorph’ to
evaluate the relationship between size and shape of
the skull averaged by species (Grafen 1989; Rohlf
2001, 2006; Martins et al. 2002; Mundry 2014);
inference is based on null-hypothesis signicance
testing (P-value). In order to present results of
regressions even in the absence of comparative
methods we included the results of standard re-
gression. Phylomorphospace was created by pro-
jecting the hippopotamid phylogeny presented in
Figure 2 onto the morphospace delimited by the
rst three PC axes.
Phylogeny
We built a synthetic phylogenetic tree (Fig.
2) using the software Mesquite 2.75 (Maddison &
Maddison 2011), based on the most recently pro-
posed phylogenetic relationships (Petronio 1986;
Boisserie 2006; Mazza & Bertini 2013) and includ-
ing all hippopotamine species considered valid. We
calibrated branch lengths in millions of years (Ma)
based on stratigraphic range in the fossil record.
The phylogeny of the family Hippopotami-
dae has been recently revised by Boisserie (2005).
The earliest representatives of the group include
the genera Archaeopotamus and Saotherium. Archaeo-
potamus occurred during the Late Miocene and it
has been considered as sister taxon to the Hippopo-
tamus-Hexaprotodon clade (Boisserie 2005). Hippopo-
tamus occurred for the rst time at least during the
Early Pliocene (Faure 1994), and it has been docu-
mented from several Pleistocene localities in Africa
(H. kaisensis, H. gorgops, H. amphibius), Europe (H.
antiquus, H. amphibius) and the Near East (H. behe-
moth) (Dietrich 1928; Hopwood 1926; Caloi et al.
1980; Capasso Barbato et al. 1982; Stuenes 1989;
Faure & Guérin 1990; Mazza 1991, 1995; Caloi
& Palombo 1994; Petronio 1995; Boisserie 2006;
Pandol & Petronio 2016). During the Pleistocene,
Hippopotamus also reached several Mediterranean
islands and Madagascar, evolving into different
dwarf species: H. pentlandi, H. melitensis, H. creutz-
burgi, H. madagascariensis, H. lemerlei, H. laloumena.
All the above-mentioned taxa seem to be derived
from H. amphibius. The phylogenetic relationships
among Pleistocene European hippopotamuses
were not investigated by Boisserie (2005) but have
been discussed in other papers (Petronio 1986,
1995; Mazza & Bertini 2013). Hexaprotodon mainly
occurred in the Indian subcontinent and south-
east Asia from the Miocene-Pliocene transition to
the late Pleistocene (Boisserie 2005 and references
therein). The earliest representative of this group
is Hex. sivalensis, while Hex. palaeindicus displays
some apomorphic features such as increased eleva-
tion of the orbits, high molar crowns and wide I/3
diameter (Boisserie 2005 and references therein).
The genus Saotherium is latest Miocene in age and
it is supposed to be the sister taxon of the extant
pigmy-hippo Choeropsis liberiensis (Boisserie 2005).
Fig. 2 - Time-calibrated phylogenetic tree of the considered species of Hippopotamidae used in this study (details are reported in methods).
Hippopotamidae skull shape 41
results
Lateral view
The rst 5 principal components of the PCA,
performed on the skulls in lateral view, summarize
78.77% of total shape variance. Figure 3A shows the
relationship between PC1 (29.84% of total shape
variance) and PC2 (18.61% of total shape variance).
Figure 3B shows the relationship between PC2 and
PC3 (13.37% of total shape variance). Figure 1S
shows the 3D plot with the relationship between
PC1, PC2 and PC3.
Positive PC1 values are associated with a mas-
sive skull having a thick and posteriorly high zygo-
matic arch, elevated orbit with respect to the cranial
roof, a concave prole of the cranial roof posterior
to the orbit, and a foramen infraorbitalis placed dis-
tant from the anterior border of the orbit. This cra-
nial shape corresponds to an Hippopotamus amphibius-
like morphology. Negative PC1 values are associated
with a shorter and less massive skull, with slender
zygomatic arch, a rather at cranial roof posterior to
the orbit, and a lower orbit relative to the cranial roof.
This shape can be observed in Choeropsis liberiensis.
At positive PC2 values the skull is massive, with
a thick zygomatic arch, a convex and downward-di-
rected dorsal prole of the neurocranial portion, and
an elevated orbit, whereas at negative PC2 values the
skull is somewhat slender, with a longer and upward-
directed neurocranial portion, a less elevated orbit
and slender zygomatic arch.
A skull that is less high, with a moderate eleva-
tion of the orbit, longer neurocranial portion, slen-
der zygomatic arch, and anteriorly projected orbit is
associated with positive PC3 values. The neurocranial
portion of the skull is shorter and the zygomatic arch
is antero-posteriorly compressed at negative PC3 val-
ues.
Variations along the three axes are shown in
Figure 4.
Specimens with Hippopotamus amphibius-like
shape are placed within the second and the third
quarters of the PC1-PC2 morphospace (Fig. 3);
Archaeopotamus and the Madagascan hippopotamids
are placed within the fourth quarter and C. liberiensis
within the rst quarter.
In the UPGMA dendrogram of cranial shape
similarities in lateral view (Fig. 5A), H. amphibius
clusters with Hex. palaeindicus and H. antiquus with
H. gorgops. The Madagascan hippopotamuses cluster
together. The Miocene A. harvardi lies close to the
Madagascan hippopotamuses. All these taxa are well
separated from the extant C. liberiensis.
A PGLS regression between shape and size of
skulls in lateral view is not signicant (P = 0.073),
nor is a standard linear regression (P = 0.07). Shape
is phylogenetically structured in the skulls in lateral
view (Fig. 3S) as revealed by the results of the phy-
signal( ) function (P < 0.005). Using the phylosig( )
function, we found that size (CS) was not phyloge-
netically structured (P = 1).
A phylogenetic heat map, plotting the input
tree and the rst ve PCs, for skulls in lateral view
is shown in Figure 6. Similarity in PC values among
species is expressed by similar color tones; the plot
summarizes the variation of PC values for each spe-
cies taking into account the phylogenetic tree.
Dorsal view
The rst 5 principal components of the PCA,
Fig. 3 - Scatterplots between PC1
and PC2 (A) and between
PC2 and PC3 (B) of the
Hippopotamidae skulls in
lateral view.
Pandol L., Martino R., Rook L. & Piras P.
42
performed on the skulls in dorsal view, summarize
82.31% of total shape variance. Figure 7A shows the
relationship between PC1 (33.63% of total shape
variance) and PC2 (20.50% of total shape variance).
Figure 7B shows the relationship between PC2 and
PC3 (12.83% of total shape variance). Figure 2S
shows the 3D plot with the relationship between
PC1, PC2 and PC3.
Positive PC1 values are associated with a rath-
er narrow skull having a roughly tapered shape, with
a zygomatic arch that is not particularly projected
laterally relatively to the orbit; the latter is slightly
oblique with respect to the antero-posterior direc-
tion. Negative PC1 values are associated with a
wider and enlarged skull, with a zygomatic arch that
is much more laterally projected with respect to the
orbit. This shape corresponds to large-sized H. am-
phibius.
At positive PC2 values the zygomatic arch
is slightly laterally projected, the orbit is anteriorly
turned a few degrees, and the neurocranium is short,
whereas at negative PC2 values the skull is longer,
and the orbit is more anteriorly placed. At positive
PC3 values, the skull is narrow, and the zygomatic
Fig. 4 - Deformation grids refer to the rst three PC axis extremes (positive and negative) for Hippopotamidae skulls in lateral view.
Fig. 5 - UPGMA dendograms for
Hippopotamidae skulls in
lateral (A) and dorsal (B)
views.
Hippopotamidae skull shape 43
arch has a rather at external prole. The skull is
wider, with an enlarged zygomatic arch at negative
PC3 values.
Variation along the three axes is shown in
Figure 8.
Hippopotamus madagascariensis and H. lemerlei
are plotted towards positive values of PC1 (Fig. 7);
C. liberiensis is plotted at negative values of PC1 and
PC2; Saotherium mingoz is plotted at negative PC1
and positive PC2 whereas H. amphibius occupies
several quarters but with larger specimens placed in
the rst quarter. Choeropsis liberiensis, H. lemerlei and
H. madagascariensis, and S. mingoz are plotted at posi-
tive values of PC3 but the rst occupies very nega-
tive values of PC2 (Fig. 7).
In the UPGMA dendrogram of cranial shape
similarities in dorsal view (Fig. 5B), H. amphibius
clusters with H. antiquus. The Madagascan hippo-
potamuses cluster together. H. gorgops lies close to
the previously mentioned two clusters. All these
taxa are well separated from the extant C. liberiensis
and from S. mingoz.
A PGLS regression between shape and
size of skulls in dorsal view is not signicant (P
= 0.414), nor is a standard linear regression (P =
0.569). Shape is phylogenetically structured for skull
Fig. 6 - A phylogenetic heat map, plotting the input tree and the rst ve PCs, for Hippopotamidae cranial shape in lateral view.
Fig. 7 - Scatterplots between PC1 and
PC2 (A) and between PC2
and PC3 (B) of Hippopotam-
idae skulls in dorsal view.
Pandol L., Martino R., Rook L. & Piras P.
44
shape in dorsal view (Fig. 4S) as revealed by the re-
sults of the physignal( ) function (P = 0.0074). Using
the phylosig( ) function, we found that size (CS) was
not phylogenetically structured (P = 0.71).
A phylogenetic heat map, plotting the input
tree and the rst ve PCs, for skulls in dorsal view is
shown in Figure 9. The similarity in PC values among
species is expressed by similar color tones; the plot
summarizes the variation of PC values for each spe-
cies taking into account the phylogenetic tree.
Fig. 8 - Deformation grids refer to the rst three PC axis extremes (positive and negative) for Hippopotamidae skulls in dorsal view.
Fig. 9 - A phylogenetic heat map, plotting the input tree and the rst ve PCs, for the Hippopotamidae cranial shape in dorsal view.
Hippopotamidae skull shape 45
dIscussIon
The Miocene Archaeopotamus
One of the oldest known hippopotamid is
the Miocene A. harvardi (= Hexaprotodon harvardi
in Coryndon 1977). This species was gracile with
unelevated orbits (Coryndon 1977; Boisserie 2005).
The main distinguishing features of the genus Ar-
chaeopotamus are canine processes poorly extended
and a highly elongate mandibular symphysis (Bois-
serie 2005). In our study the species A. harvardi is
represented by one specimen that displays negative
values of PC1 and PC2 and positive values of PC3
in lateral view. These values are associated with a
shorter and less massive skull, low orbits and a slen-
der zygomatic arch. The latter features are also pres-
ent in the Madagascan hippos.
Continental Hippopotamus and Hexaprotodon
Elevated orbits, a long facial region and a
short postorbital part of the skull in Hippopotami-
dae are related to specialization for a semiaquatic
lifestyle (Coryndon 1977). Hippopotamus gorgops and
H. antiquus share these features, and probably were
more aquatic than H. amphibius (Stuenes 1989). Ac-
cording to Mazza (1991) H. antiquus from Upper
Valdarno is morphologically similar to H. gorgops of
Olduvai Bed II. In our analysis, H. gorgops and H.
antiquus display similar shapes, with negative values
of PC2 and positive values of PC1. The similarity
in shape of these two species is reinforced by the
UPGMA analysis of skull shape acquired in lateral
view.
The adaptation to a semiaquatic life style
evolved independently in Hippopotamus and Hexa-
protodon suggesting a convergence between the two
lineages, as previously discussed in several studies
(Boisserie 2005; Boisserie et al. 2007, 2011 and ref-
erences therein).
In the present analysis, Hex. palaeindicus falls
within the morphospace occupied by H. amphibius,
suggesting similar shape of the skull. Six incisors
have been used in the past as a distinguishing charac-
ter in Hexaprotodon (Coryndon 1977) but this genus
also differs from Hippopotamus by having a very high
robust mandibular symphysis and canine processes
that are not laterally extended (Boisserie 2005). In
addition, Hexaprotodon is identied by low-crowned
molars, while Hippopotamus has high-crowned ones
(Weston 2000). Hexaprotodon is also characterized by
slender and less massive postcranial remains, sug-
gesting that it was less well adapted for walking on
mud (Weston 2000). Boisserie (2005) recognized an
evolutionary trend within the genus Hexaprotodon
with increase in orbit elevation and increase in mo-
lar crown height.
Insular hippopotamuses
Insular hippopotamids are characterized by
reduction in size with respect to their continental
ancestors, by a general decrease of the height of
the orbits (Caloi & Palombo 1994) and by brain
size reduction (Weston & Lister 2009). According
to van der Geer et al. (2018) the largest amount of
morphological variation in dwarfed hippos is in
muzzle shape, which becomes anteriorly low. This
new arrangement of the anterior skull requires a re-
structuring of the dental battery that in some cases
even includes the loss of a premolar (van der Geer
et al. 2018). In addition, some works suggest that
insular hippopotamids show an increasing trend to-
ward terrestrialization (Boisserie et al. 2011). This
hypothesis is based on three main morphological
comparisons: limb, cranial and tooth morphology.
Limbs become shorter, more erect and with re-
stricted lateral movement (Houtemaker & Sondaar
1979). This new structure of the limbs is linked to
a ‘low gear locomotion’ adaptation to the different
rocky grounds of the Mediterranean islands con-
quered by the insular species (Sondaar 1977).
Hippopotamus pentlandi from Sicily and Malta
shows a slight reduction in size with respect to its
continental ancestor H. amphibius and more robust
limb bones (Boisserie 2005). It is unclear if limb
robustness can provide information on water de-
pendence (Boisserie et al. 2011). Hippopotamus pent-
landi displays a shorter muzzle, less developed nasal
region and a wider diastema C/P2 relative to H. an-
tiquus (Caloi & Palombo 1986). The shape of the
Sicilian hippopotamus still resembles that of H. am-
phibius in lateral view and falls well within the mor-
phospace occupied by the extant species, suggesting
that its adaptation to the insular domain was not yet
fully attained. Unfortunately, the shape analysis of
H. pentlandi is not exhaustive; the remains attributed
to this species are not well preserved and only one
skull was well enough preserved to be included in
our study.
The dwarfed hippos of Madagascar exhibit
different degrees of muzzle shortening. Hippopota-
Pandol L., Martino R., Rook L. & Piras P.
46
mus madagascariensis is characterized by a moderately
short muzzle, whereas H. lemerlei does not exhibit
shortening but has an anterior narrowing instead.
The results obtained by van der Geer et al. (2018)
that mainly focus on the differences in the anterior
portion of the skull reinforce the idea that dwarfed
hippos are not ‘downscaled mainland hippos’.
Differences in cranio-dental morphology of
the Madagascan species, such as the different orbit
heights (Stuenes 1989), suggest that they occupied
different ecological niches (Rakotovao et al. 2014),
with H. madagascariensis being more terrestrial than
H. lemerlei. Hippopotamus madagascariensis presents a
different orbit orientation, with orbits more elevat-
ed than H. lemerlei, and it has smaller dimensions
relative to H. amphibius (Rakotovao et al. 2014). The
skull of H. madagascariensis is robust and character-
ized by a facial prolongation with a thin supraorbital
margin, a short postorbital part, and orbits as high
as they are wide (Stuenes 1989). Despite the adapta-
tion to terrestriality, cranial shape in the Madagas-
can hippos seems not to be closely related to the
cranial shape of C. liberiensis. These species can be
found at negative values of PC1, but the morpho-
space occupied is different, supporting the distance
between the hippos of Madagascar and C. liberiensis.
The UPGMA analysis supports the closeness of H.
madascariensis and H. lemerlei.
The skull morphology of the Malagasy hip-
pos displays features typical of the genus Hexapro-
todon (Stuenes 1989). These features are a double
rooted rst premolar and the tip-to-tip occlusion in
H. madascariensis, while, regarding both species, the
greatest breadth of the nasals falls within the varia-
tion typical of Hexaprotodon more than that of Hip-
popotamus. In our study the Madagascan species fall
in the same morphospace as Hippopotamus at posi-
tive values of PC1 and negative values of PC2 in
lateral view, and at negative values of PC2 and PC3
for dorsal view. The analyses here performed do not
highlight similarities between the hippos from Mad-
agascar and the genus Hexaprotodon (albeit the latter
group is here represented by very few specimens).
The extant pygmy hippopotamus
It has been suggested that C. liberiensis is
a dwarfed version of the common hippo (Gould
1975). It is tempting to link this hypothesis to the
forest refugia hypothesis (Mayr & O’Hara 1986),
stipulating major contractions of forest habitats
into small areas during past episodes of aridity. Cho-
eropsis liberiensis could thus have evolved an insular
morphology in restricted forest patches. It displays
some features common to island hippopotamid
morphology, including reduced transverse move-
ment of the front limb (Houtemaker & Sondaar
1979) and a somewhat lophodont cheek dentition,
but these are not sufcient to support a dwarf-
ing event. It has been clearly demonstrated that
C. liberiensis is not a dwarfed Hippopotamus (Weston
2003a), but rather evolved as a separate lineage, dis-
tinct from all other hippopotamids, since the latest
Miocene (Boisserie 2005), and its size could be a
plesiomorphic trait, in contrast to Mediterranean
dwarf species that evolved from larger relatives.
The genus names Choeropsis and Hexaprotodon
appear interchangeably in the scientic literature,
although Boisserie (2005) concluded that the com-
bination of primitive and derived characteristics
of the extant pygmy hippo place it in a distinct lin-
eage, validating the genus Choeropsis, and restricting
the genus Hexaprotodon to the fossil lineage mostly
found in Asia. According to Weston (2000) the pyg-
my hippo was labelled as a ‘living fossil’ because it
shares more traits with extinct ancestral clades than
with H. amphibius. Most of its cranial traits are ple-
siomorphic, while the mandible and the dentition
display some very derived features. The main ple-
siomorphic traits in Choeropsis are a weak extension
of the canine processes, a slender zygomatic arch in
ventral view, a lachrymal separated from the nasal
by a long maxillary process of the frontal, and elon-
gated and transversally rounded braincase (Boisserie
2005). In addition, the pigmy hippo shows a down-
turned sagittal crest, a feature that is generally re-
garded as a plesiomorphy, reinforcing the ‘primitive
aspect’ of the Liberian hippos (Boisserie 2005). The
theory of the ‘different lineage’ of Weston (2003) is
supported also by the work of van der Geer et al.
(2018): the mainland dwarf hippo Choeropsis is not
a downscaled version of the species Hippopotamus.
The cranial shape of the considered taxa revealed
that C. liberiensis does not t well within insular hip-
pos derived from H. amphibius and it occurs in dif-
ferent areas of the morphospace.
The genus Saotherium
Another genus that shows a mosaic of char-
acters is Saotherium. According to Boisserie (2003)
the species S. mingoz, from the Mio-Pliocene bound-
Hippopotamidae skull shape 47
ary, combines features generally considered derived
(e.g., a short premolar row) and primitive charac-
ters (e.g., low orbits and a weak development of ca-
nine processes). Another important feature of this
species is the relative height of the skull above the
molars. These characteristics are unique in hippo-
potamids and the introduction of the new genus
Saotherium, replacing Hexaprotodon, is therefore jus-
tied (Boisserie 2003). Parsimony analysis of Hip-
popotamidae relates Saotherium with the genus Choe-
ropsis (Boisserie 2005). The Liberian hippos lack the
cranial structure typical of Saotherium and therefore
these two species share features that are mainly ple-
siomorphic or convergent (Boisserie 2005).
Boisserie (2007) gave a new interpretation of
the mosaic features of Choeropsis, this species pres-
ents physiological adaptations to semiaquatic envi-
ronments and some of the archaic features of its
skull could therefore be interpreted as a secondary
adaptation to the shaded rainforest. This adapta-
tion would explain the low position of the orbits
and the sagittal crest on the cranium of Choeropsis,
which could be useful for a better penetration of
the dense rainforest vegetation. This feature is also
present in Saotherium (Boisserie 2005) and it could
be interpreted as a link between these two species.
This hypothesis may be reinforced by the UPGMA
analysis on dorsal view data, where these two spe-
cies are closely related. However, the relationships
between Choeropsis and Saotherium is still difcult to
understand.
conclusIons
The present paper represents a contribution
towards the understanding of the variation in skull
shape in Hippopotamidae by means of shape analy-
sis (Geometric Morphometrics) and modern statis-
tical approaches. Two-dimensional modern shape
analysis applied to dorsal and lateral views of extant
and extinct Hippopotamidae species highlighted
several points that will be further investigated in fu-
ture works.
The skull shape of the pigmy hippo doesn’t
match those of insular hippopotamuses; the
different specimens are plotted in different areas
of the morphospace. Choeropsis liberiensis cannot
be considered as a model for insular fossil species
and the appearance of terrestrial characters within
Hippopotamus evolved independently following
a different morphological pattern that needs
to be further investigated (Fig. 10). Similarities
between Hexaprotodon and Hippopotamus have been
detected but need further investigation, as well as
the similarities between H. gorgops and H. antiquus.
Despite the morphological convergences within the
group, cranial shape (for the chosen congurations)
in Hippopotamidae is phylogenetically structured
while this does not hold for size. Small-sized hippos
occur within different lineages but are related to
different cranial shapes.
Acknowledgments: We thank two anonymous reviewers for
their suggestions and comments. We are also grateful to the Editor L.
Werdelin. L.P. thanks E. Cioppi (IGF), M.C. De Angelis (MPLBP),
R. Manni and C. Petronio (MPUR), and R. Carlini (MZR) for their
help and assistance during his visits to the hippopotamid fossil col-
lections. This paper has been developed within the research project
“Ecomorphology of fossil and extant Hippopotamids and Rhinoc-
erotids” granted to L.P. by the University of Florence (“Progetto
Giovani Ricercatori Protagonisti” initiative). Author contributions:
L.P. designed the study; L.P. and R.M. collected the data; L.P. and
P.P. performed the analyses; all the authors wrote data interpretation
and discussion.
RefeRences
Adams D.C., Rohlf F.J. & Slice D.E. (2004) - Geometric mor-
phometrics: ten years of progress following the ‘revolu-
tion’. Italian Journal of Zoology, 71(1): 5-16.
Adams D.C., Collyer M.L. & Kaliontzopoulou A. (2019) -
Geomorph: Software for geometric morphometric
Fig. 10 - Head life appearance comparing Choeropsis liberiensis and
Hippopotamus amphibius as reconstructed by Edoardo Pre-
viato (Msc student in Natural Sciences, Sapienza University
of Rome).
Pandol L., Martino R., Rook L. & Piras P.
48
analyses. R package version 3.1.0. https://cran.r-project.
org/package=geomorph.
Boisserie J. R. (2005) - The phylogeny and taxonomy of Hip-
popotamidae (Mammalia: Artiodactyla): a review based
on morphology and cladistic analysis. Zoological Journal of
the Linnean Society, 143(1): 1-26.
Boisserie J. R. (2006) - Langebaanweg: a unique window on
early hippopotamid evolution in southern Africa. Afri-
can Natural History, 2: 181-182.
Boisserie J.R. (2007) - Late Miocene Hippopotamidae from
Lemudong’o, Kenya. Kirtlandia, 56: 158-162.
Boisserie J.R., Fisher R.E., Lihoreau F. & Weston E.M. (2011)
- Evolving between land and water: key questions on the
emergence and history of the Hippopotamidae (Hip-
popotamoidea, Cetancodonta, Cetartiodactyla). Biological
Review, 86(3): 601-625.
Bookstein F.L. (1986) - Size and shape spaces for landmark
data in two dimensions. Statistical Science, 1(2): 181-222.
Bookstein F.L. (1991) - Morphometric tools for landmark
data. Cambridge University Press, Cambridge, 435 pp.
Caloi L., Palombo M.R. & Petronio C. (1980) - Resti cranici di
Hippopotamus antiquus (= H. major) e Hippopotamus amphi-
bius conservati nel Museo di Paleontologia dell’Università
di Roma. Geologica Romana, 19: 91-119.
Caloi L. & Palombo M.R. (1986) - Some remarks about bio-
metric characters of skulls of Hippopotamus pentlandi Von
Meyer (Cannita Cave, Sicily) and pleistocene hippopot-
ami of Western Europe/Osservazioni sui caratteri bio-
metrici del cranio di Hippopotamus pentlandi Von Meyer
della grotta della Cannita (Sicilia) e degli ippopotami
pleistocenici dell’Europa Occidentale. Hystrix, Italian
Journal of Mammology, 1(2): 137-159.
Caloi L. & Palombo M.R. (1994) - Le faune a grandi mam-
miferi del Pleistocene superiore dell’Italia centrale: bio-
stratigraa e paleoambiente. Bollettino del Servizio Geologico
d'Italia, 111(7): 503-514.
Capasso Barbato L., Kotsakis T. & Petronio C. (1982) - Nuo-
vi resti e osservazioni logenetiche su Hippopotamus
creutzburgi Boekschoten and Sondaar (Hippopotamidae,
Mammalia) del Pleistocene di Creta (Grecia). Geologica
Romana, 21: 61-77.
Coryndon S.C. (1977) - The taxonomy and nomenclature
of the Hippopotamidae (Mammalia, Artiodactyla) and
a description of two new fossils species. Proceedings of
the Koninklijke Nederlandse Akademie van Wetenschappen,
B(80): 61-88.
Coughlin B. & Fish. F. (2009) - Hippopotamus underwater lo-
comotion: Reduced gravity movements for a massive
mammal. Journal of Mammology, 90(3): 675-679.
Dietrich W.O. (1928) - Pleistocäne deutsch ostafrikanische
Hippopotamus-reste. Wissenschaftliche Ergebnisse des Oldoway
Expedetion Herausgeben Von Prof. Dr. Reck, 1913: 3-41.
Faure M. (1994) - Les Hippopotamidae (Mammalia, Artiodac-
tyla) du rift occidental (bassin du lac Albert, Ouganda).
Etude Préliminaire. In: Senut B. & Pickford M. (Eds)
- Geology and paleobiology of the Albertine Rift Val-
ley, Uganda-Zaïre; II, Paleobiology: 321-337. CIFEG,
Orléans.
Faure M. & Guérin C. (1990) - Hippopotamus laloumena nov. sp.,
la troisième espèce d’hippopotame Holocène de Mada-
gascar. Comptes Rendus de l’Académie des Sciences, Paris, série
310(2): 1299-1305.
Goodall C. (1991) - Procrustes methods in the statistical analy-
sis of shape. Journal of the Royal Statistical Society. Series B
(Methodological), 53(2): 285-339.
Gould S.J. (1975) - On the scaling of tooth size in mammals.
American Zoologist, 15(2): 353-362.
Grafen A. (1989) - The phylogenetic regression. Philosophical
Transaction of the Royal Society of London B, 326(1233): 119-
157.
Gunz P. & Mittoeroecker P. (2013) - Semilandmarks: a method
for quantifying curves and surfaces. Hystrix, Italian Jour-
nal of Mammology, 24(1): 103-109.
Hopwood A.T. (1926) - Some mammalia from the Pliocene of
Homa Mountain, Victoria Nyanza. Annals and Magazine
of Natural History, Series, 9(18): 266-272.
Houtemaker J. & Sondaar P.Y. (1979) - Osteology of the fore-
limb of the Pleistocene dwarf hippopotamus from Cy-
prus with special reference to phylogeny and function.
Proceedings Koninklijke Nederlande Akademie van Wefenschap-
pen , B, 82: 411-448.
Márquez E.J., Cabeen R., Woods R.P. & Houle D. (2012) - The
measurement of local variation in shape. Evolutionary bi-
ology, 39(3): 419-439.
Martins E.P., Diniz-Filho J.A. & Housworth E.A. (2002) -
Adaptive constraints and the phylogenetic comparative
method: a computer simulation test. Evolution, 56(1):
1-13.
Mayr E. & O’Hara R.J. (1986) - The biogeographic evidence
supporting the Pleistocene forest refuge hypothesis. Evo-
lution, 40(1): 55-67.
Mazza P. (1991) - Interrelation between Pleistocene hippopo-
tami of Europe and Africa. Bollettino della Società Paleonto-
logica Italiana, 30(2): 153-186.
Mazza P. & Bertini A. (2013) -Were Pleistocene hippopota-
muses exposed to climate-driven body size changes?
Boreas, 42(1): 194-209.
Maiorino L., Farke A.A., Kotsakis T. & Piras P. (2013) - Is To-
rosaurus Triceratops? Geometric morphometric evidence
of late Maastrichtian ceratopsid dinosaurs. PLOS ONE,
8, e81608.
Mundry R. (2014) - Statistical issues and assumptions of phy-
logenetic generalized least squares. In: Modern phylo-
genetic comparative methods and their application in
evolutionary biology: 131-153. Berlin, Garamszegi LZ,
editor.
Perez S.I., Bernal V. & Gonzalez P.N. (2006) - Differences
between sliding semi-landmark methods in geometric
morphometrics, with an application to human cranio-
facial and dental variation. Journal of Anatomy, 208(6):
769-784.
Pandol L. & Maiorino L. (2016) - Reassessment of the
largest Pleistocene rhinocerotine Rhinoceros platyrhinus
(Mammalia, Rhinocerotidae) from the Upper Siwaliks
(Siwalik Hills, India). Journal of Vertebrate Paleontology, 36
(2): e1071266.
Hippopotamidae skull shape 49
Pandol L. & Petronio C. (2015) - A brief review of the oc-
currences of Pleistocene Hippopotamus (Mammalia, Hip-
popotamidae) in Italy. Geologia Croatica, 68(3): 313-319.
Pandol L., Marra A.C., Carone G., Maiorino L. & Rook L.
(2019) - A new rhinocerotid (Mammalia, Rhinocerotid-
ae) from the latest Miocene of Southern Italy. Historical
Biology, doi: 10.1080/08912963.2019.1602615.
Petronio C. (1986) - Nuovi resti di ippopotamo del Pleisto-
cene medio-inferiore dei dintorni di Roma e problemi di
tassonomia e logenesi del gruppo. Geologica Romana, 25:
63-73.
Petronio C. (1995) - Note on the taxonomy of Pleistocene
hippopotamuses. Ibex, 3: 53-55.
Piras P., Teresi L., Buscalioni A.D. & Cubo J. (2009) - The
shadow of forgotten ancestors differently constrains the
fate of Alligatoroidea and Crocodyloidea. Global Ecology
and Biogeography, 18: 30-40.
Piras P., Marcolini F., Raia P., Curcio M.T. & Kotsakis T. (2010)
- Ecophenotypic variation and phylogenetic inheritance
in rst lower molar shape of extant Italian populations
of Microtus (Terricola) savii (Rodentia). Biological Journal of
the Linnean Society of London 99: 632-647.
Piras P., Buscalioni A.D., Teresi L., Raia P., Sansalone G., Kot-
sakis T. & Cubo J. (2014) - Morphological integration
and functional modularity in the crocodilian skull. Inte-
grative Zoology, 9: 498-516.
Rakotovao M., Lignereux Y., Orliac M.J., Duranthon F. & An-
toine P.O. (2014) - Hippopotamus lemerlei Grandidier, 1868
et Hippopotamus madagascariensis Guldberg, 1883 (Mam-
malia, Hippopotamidae): anatomie crânio-dentaire et
revision systématique. Geodiversitas, 36(1): 117-161.
Revell L.J. (2012) - Phytools: an R package for phylogenetic
comparative biology (and other things). Methods in Ecol-
ogy and Evolution, 3(2): 217-223.
Rohlf F.J. (2001) - Comparative methods for the analysis of
continuous variables: geometric interpretations. Evolu-
tion, 55(11): 2143-2160.
Rohlf F.J. (2006) - TpsDig, program for Digitizing landmarks
and outline for Geometric morphometric analyses, Ver-
sion 2.04. Department of Ecology and Evolution, State
University of New York at Stony Brook.
Rohlf F.J. (2013) - TpsDIG, version 2.17. SUNY Stonybrook.
Schlager S. (2014) - Morpho: Calculations and visualisations
related to Geometric Morphometrics. R-package ver-
sion 2.0. 3-1.
Sondaar P.Y. (1977) - Insularity and its effect on mammal evo-
lution. Major patterns in vertebrate evolution: 671-707.
Springer, Boston.
Spaan A. (1996) - Hippopotamus creutzburgi: the case of the Cre-
tan Hippopotamus. Pleistocene and Holocene fauna of
Crete and its rst settlers. World Archaeology, 28: 99-111.
Stevenson-Hamilton J. (1912) - Animal Life in Africa. E.P.
Dutton and Company, New York, 539 pp.
Stuenes S. (1989) - Taxonomy, habits, and relationships of the
subfossil Madagascan hippopotami Hippopotamus lemerlei
and H. madagascariensis. Journal of Vertebrate Paleontology, 9
(3): 241-268.
van der Geer A.A.E., Lyras G.A., Mitteroecker P. & MacPhee
R.D. (2018) - From Jumbo to Dumbo: Cranial Shape
Changes in Elephants and Hippos During Phyletic
Dwarng. Evolutionary Biology, 45(3): 303-317.
Weston E.M. (2000) - A new species of hippopotamus Hexa-
protodon lothagamensis (Mammalia: Hippopotamidae) from
the late Miocene of Kenya. Journal of Vertebrate Paleontol-
ogy, 20: 177-185.
Weston E.M. (2003) - Evolution of ontogeny in the hippo-
potamus skull: using allometry to dissect developmental
change. Biological Journal of the Linnean Society, 80(4): 625-
638.
Weston E.M. & Lister A. (2009) - Insular dwarsm in hippos
and a model for brain size reduction in Homo oresiensis.
Nature, 459(7243): 85-88.
... Moreover, the skull is shorter and less massive than in the extant hippo, with lower orbits and a slender zygomatic arch. The latter features are also present in the pigmy-insular Madagascan hippos (Pandolfi et al., 2020), which are characterized by reduction of size with respect to their continental ancestors, by a general decrease of the height of the orbits (Caloi and Palombo, 1994), and by brain size reduction (Weston and Lister, 2009). The opposite condition can be found in more derived Hexaprotodon (H. ...
... The opposite condition can be found in more derived Hexaprotodon (H. palaeindicus) and in H. amphibius (Boisserie, 2005(Boisserie, , 2007Boisserie et al., 2011;Pandolfi et al., 2020): in these species, orbits are elevated beyond the cranial roof. Elevated orbits, a long facial region, and a short postorbital part of the skull in Hippopotamidae are related to specialization for a semiaquatic lifestyle. ...
... Semi-landmarks were used to capture the morphology of complex outlines where anatomical homology is difficult to recognize. See Figure 1 in Pandolfi et al. (2020) for landmarks definition. Semi-landmarks assume that curves or contours are homologous among specimens (Adams et al., 2004;Perez et al., 2006). ...
Article
Full-text available
In modern shape analysis, deformation is quantified in different ways depending on the algorithms used and on the scale at which it is evaluated. While global affine and non-affine deformation components can be decoupled and computed using a variety of methods, the very local deformation can be considered, infinitesimally, as an affine deformation. The deformation gradient tensor F can be computed locally using a direct calculation by exploiting triangulation or tetrahedralization structures or by locally evaluating the first derivative of an appropriate interpolation function mapping the global deformation from the undeformed to the deformed state. A suitable function is represented by the thin plate spline (TPS) that separates affine from non-affine deformation components. F, also known as Jacobian matrix, encodes both the locally affine deformation and local rotation. This implies that it should be used for visualizing primary strain directions (PSDs) and deformation ellipses and ellipsoids on the target configuration. Using C = F T F allows, instead, one to compute PSD and to visualize them on the source configuration. Moreover, C allows the computation of the strain energy that can be evaluated and mapped locally at any point of a body using an interpolation function. In addition, it is possible, by exploiting the second-order Jacobian, to calculate the amount of the non-affine deformation in the neighborhood of the evaluation point by computing the body bending energy density encoded in the deformation. In this contribution, we present (i) the main computational methods for evaluating local deformation metrics, (ii) a number of different strategies to visualize them on both undeformed and deformed configurations, and (iii) the potential pitfalls in ignoring the actual three-dimensional nature of F when it is evaluated along a surface identified by a triangulation in three dimensions.
... Further diagnostic characters of H. antiquus cranial and postcranial elements were described in and Mazza (1995). According to Mazza (1991Mazza ( , 1995 and Pandolfi et al. (2020), the skull of H. antiquus resembles that of the African species H. gorgops. In particular, Mazza (1991) claimed that H. antiquus is apparently closer to H. gorgops collected from Olduvai Beds I-II. ...
... The orbits in H. pentlandi are more elevated than in H. amphibius and the mandible displayed a particularly low condyle (Caloi and Palombo 1983). According to Pandolfi et al. (2020) the cranial shape in lateral view of the Sicilian hippopotamus still resembles that of H. amphibius, and a full adaptation to the insular environment was not yet attained. For additional diagnostic characters of H. pentlandi cranial and postcranial elements refer to Accordi (1955), Palombo (1983, 1986), and Capasso Barbato and Petronio (1983b). ...
Article
The genus Hippopotamus dispersed from Africa to Europe during the Pleistocene and the taxon is testified from different Italian localities. Despite the large number of available specimens, Hippopotamus is mainly documented by fragmentary remains and detailed stratigraphic and/or geographic data are often absent or uncertain. In Italy, hippopotamuses are recorded in at least 100 fossiliferous localities spanning from the late Early Pleistocene to the Late Pleistocene. At present, three different Pleistocene species are recognised: H. antiquus from Early to Middle Pleistocene, H. amphibius from Middle to Late Pleistocene and H. pentlandi from late Middle-early Late Pleistocene Sicilian localities. Upper Valdarno basin (Tuscany), Chiusi basin (Tuscany), Colfiorito basin (Umbria), Sant’Oreste (Latium), Puntali Cave (Sicily) and Acquedolci (Sicily) are just a few fossiliferous Italian sites where significative findings of Hippopotamus were collected during the 19th and 20th centuries.
... The possibility of the presence of three species of hippos in Europe during the Pleistocene was exposed by Mazza (1991), although later works ruled out this possibility due to the absence of significant differences in the morphology among these taxa (Petronio, 1995). In turn, the similarities between H. antiquus (in the broad sense) and H. gorgops have been highlighted, raising the possibility of considering H. gorgops a junior subjective synonym of H. antiquus (Martínez-Navarro et al., 2015;Pandolfi et al., 2020). ...
Article
Full-text available
The dispersal of the genus Hippopotamus from Africa to Europe in the Early Pleistocene has been widely discussed in the last decades. Most authors propose a single entry of forms similar to the African species Hippopotamus gorgops, named H. antiquus upon entering Europe. However, other authors propose a first entry of forms similar to the African taxon H. kaisensis, named H. antiquus upon entering Europe, and an Early Pleistocene second entry of forms similar to H. gorgops, named H. tiberinus upon entering Europe. In this work we evaluated the current state of the debate reviewing the fossil record of genus Hippopotamus highlighting the last findings in sites from central Africa to Europe and their possible passageways.
Article
Full-text available
The classification of most mammalian orders and families is under debate and the number of species is likely greater than currently recognized. Improving taxonomic knowledge is crucial, as biodiversity is in rapid decline. Morphology is a source of taxonomic knowledge, and geometric morphometrics applied to two dimensional (2D) photographs of anatomical structures is commonly employed for quantifying differences within and among lineages. Photographs are informative, easy to obtain, and low cost. 2D analyses, however, introduce a large source of measurement error when applied to crania and other highly three dimensional (3D) structures. To explore the potential of 2D analyses for assessing taxonomic diversity, we use patas monkeys (Erythrocebus), a genus of large, semi‐terrestrial, African guenons, as a case study. By applying a range of tests to compare ventral views of adult crania measured both in 2D and 3D, we show that, despite inaccuracies accounting for up to ¼th of individual shape differences, results in 2D almost perfectly mirror those in 3D. This apparent paradox might be explained by the small strength of covariation in the component of shape variance related to measurement error. A rigorous standardization of photographic settings and the choice of almost coplanar landmarks are likely to further improve the correspondence of 2D to 3D shapes. 2D geometric morphometrics is, thus, appropriate for taxonomic comparisons of patas ventral crania. Although it is early to generalize, our results corroborate similar findings from previous research in mammals, and suggest that 2D shape analyses are an effective heuristic tool for morphological investigation of small differences. This article is protected by copyright. All rights reserved.
Article
Full-text available
Free online view at https://rdcu.be/KMNX Members of the mammalian families Elephantidae and Hippopotamidae (extant and extinct elephants and hippos) include extinct dwarf species that display up to 98% decrease in body size compared to probable ancestral sources. In addition to differences in body mass, skulls of these species consistently display distinctive morphological changes, including major reduction of pneumatised areas in dwarf elephants and shortened muzzles in dwarf hippos. Here we build on previous studies of island dwarf species by conducting a geometric morphometric analysis of skull morphology and allometry in target taxa, living and extinct, and elaborate on the relation between skull size and body size. Our analysis indicates that skull size and body size within terrestrial placental mammals scale almost isometrically (PGLS major axis slope 0.906). Furthermore, skull shape in dwarf species differed from both their ancestors and the juveniles of extant species. In insular dwarf hippos, the skull was subject to considerable anatomical reorganisation in response to distinct selection pressures affecting early ontogeny (the “island syndrome”). By contrast, skull shape in adult insular dwarf elephants can be explained well by allometric effects; selection on size may thus have been the main driver of skull shape in dwarf elephants. We suggest that a tightly constrained growth trajectory, without major anatomical reorganization of the skull, allowed for flexible adaptations to changing environments and was one of the factors underlying the evolutionary success of insular dwarf elephants.
Article
Full-text available
Several Pleistocene hippopotamid remains have been collected from the Italian Peninsula, but some of them lack any stratigraphic data or details about the fossil localities. The Hippopotamus was present in Italy and Western Europe from the latest Villafranchian (late Early Pleistocene), whereas the oldest records of the genus (Coste San Giacomo and Chiusi basin) need to be confirmed by further evidence. Hippopotamus antiquus is recorded from Italy until MIS 15, whereas H. amphibius is first known from the late Middle Pleistocene (MIS 13 or 11) and survived in Southern Italy until MIS 4/3.
Article
A new species of Rhinocerotidae (Perissodactyla), ‘Ceratotherium’ advenientis sp. nov., from the late Miocene (8.1–7.2 Ma) locality of Cava Gentile, Calabria (Southern Italy), is described. ‘Ceratotherium’ advenientis displays morphological characters close to Rhinocerotina, in particular to dicerotines, and can be distinguished from the late Miocene elasmotheres, teleoceratines and aceratheres recorded in Eurasia and Africa. The new taxon clearly differs from the European latest Miocene species Dihoplus schleiermacheri, Dihoplus pikermiensis, ‘Dihoplus’ megarhinus and Ceratotherium neumayri, and from the African species Ceratotherium douariense, Ceratotherium? primaevum and Paradiceros mukirii. ‘Ceratotherium’ advenientis also differs from the Chinese dicerotine Diceros gansuensis and from the extant African species. The new taxon is characterized by peculiar features, in particular in the morphology and dimension of the neurocranial portion, and by having a nuchal crest wider than in the extant African rhinoceroses, C. neumayri, C. douariense, and European latest Miocene species. A cladistic analysis places ‘Ceratotherium’ advenientis in a polytomy with the extant Diceros bicornis, C. neumayri and a small clade composed by C. simum and C. antiquitatis. The African affinities of the new taxon support the Calabrian-Peloritan arc as a northern extension of the African continental shelf during the late Miocene.
Article
The prevailing explanation for the observed distributional patterns and areas of endemism of tropical forest organisms is the Pleistocene refuge hypothesis, which proposes that wide-ranging ancestral taxa were isolated into forest refuges during certain glacial periods, and that this isolation provided them with the opportunity to speciate. John Endler has recently argued that two predictions of the refuge hypothesis-that contact zones between vicars should be between refuges and that contact zones of rapidly reproducing butterflies should be wider than those of more slowly reproducing birds-are not borne out by the evidence. Endler therefore rejects the refuge hypothesis. We show that the data available are far too imprecise to permit any conclusions regarding contact zone widths and that, according to our reanalysis of the African bird data used by Endler, all the contact zones between vicars do indeed occur between refuges, exactly where they are expected. Additional strong support for the refuge hypothesis comes from the existence of many taxa endemic to the particular forest areas which have been postulated as refuges and from fragmented taxa which are still allopatric, never having come into secondary contact.
Article
eprint link http://www.tandfonline.com/eprint/RMd2IkpggRyTkimPiU8b/full We describe and figure a well-preserved, large skull of a rhinoceros, NHMUK 36661, collected in 1860 from Upper Siwalik deposits. This specimen can be referred to Rhinoceros platyrhinus. Comparison with the type material of R. platyrhinus revealed that several specimens previously referred to this taxon, including the lectotype, should instead be assigned to Rhinoceros sp. (potentially R. sivalensis or R. unicornis). Therefore, we here provide new detailed cranial and dental characters for R. platyrhinus, which is currently known only by a few specimens collected from a restricted area of northern India. We suggest that the generic name Punjabitherium erected for R. platyrhinus represents a junior synonym of Rhinoceros due to the morphological affinities of NHMUK 36661 with R. unicornis. A principal component analysis and a cluster analysis confirmed the morphological similarities between R. platyrhinus and R. unicornis. Rhinoceros platyrhinus represents the largest rhinocerotine species in Eurasia and is characterized by a long skull and high-crowned teeth, suggesting that it was a grazer rather than a mixed feeder such as R. unicornis. This is supported by a cluster analysis on the upper teeth. The progressive increase in aridity from ca. 12 Ma to Recent in northern India could have affected the dietary regime of R. platyrhinus towards to a more grazer-like diet.
Article
An hippopotamus mandible found on the East coast of Madagascar, near Mananjary, was published in 1922 as a new variety of the recent African hippo, Hippopotamus amphibius var. standini Monnier and Lamberton. The taxon is a nomen nudum but the mandible clearly pertains to a new species, owing to its size and proportions, and we name it H. laloumena. We attribute to it five metapodials from the old collections of the Malagasy Academy. A third hippo species exists therefore in Madagascar during the Holocene, together with the very common small H. lemerlei and the much rarer H. madagascariensis nearly as tall but differently proportioned. There is an abridged English version. -English summary
Chapter
Using phylogenetic generalized least squares (PGLS) means to fit a linear regression aiming to investigate the impact of one or several predictor variables on a single response variable while controlling for potential phylogenetic signal in the response (and, hence, non-independence of the residuals). The key difference between PGLS and standard (multiple) regression is that PGLS allows us to control for residuals being potentially non-independent due to the phylogenetic history of the taxa investigated. While the assumptions of PGLS regarding the underlying processes of evolution and the correlation of the predictor and response variables with the phylogeny have received considerable attention, much less focus has been put on the checks of model reliability and stability commonly used in case of standard general linear models. However, several of these checks could be similarly applied in the context of PGLS as well. Here, I describe how such checks of model stability and reliability could be applied in the context of a PGLS and what could be done in case they reveal potential problems. Besides treating general questions regarding the conceptual and technical validity of the model, I consider issues regarding the sample size, collinearity among the predictors, the distribution of the predictors and the residuals, model stability, and drawing inference based on P-values. Finally, I emphasize the need for reporting checks of assumptions (and their results) in publications.